论文部分内容阅读
主要研究遥感湖泊面积亚像元分解提取方法和空间尺度效应,为遥感湖泊面积提取、检验及基于此的局地气候变化分析提供科学的基础数据。在对TM遥感数据进行升尺度处理的基础上,采用混合调制匹配滤波(Mixture Tuned Matched Filtering,MTMF)进行亚像元分解,得到不同空间分辨率的湖泊面积。进而分析不同面积湖泊随遥感空间尺度的变化。结果表明:(1)当通过对高空间分辨率的遥感数据重采样获取多尺度遥感影像进行湖泊面积提取及湖泊空间尺度效应分析时,采用最近邻法比像元聚合重采样法更合理。(2)MTMF亚像元分解法可以用于基于水体光谱特征的遥感湖泊边界提取和面积计算,但边界提取过程中容易将湖泊与河流或其他非湖泊的水体混淆。(3)遥感湖泊面积的提取结果受所用遥感影像空间分辨率的影响较大,影像的空间分辨率越低,湖泊面积提取的偏差越大,尤其对面积较小的湖泊。
The main research is remote sensing lake area sub-pixel decomposition extraction method and spatial scale effect for the remote sensing lake area extraction, testing and based on the analysis of local climate change to provide scientific and basic data. Based on the scale-up of TM remote sensing data, sub-pixel decomposition is performed by using Mixture Tuned Matched Filtering (MTMF) to obtain the lake area with different spatial resolution. Then the change of the spatial scale of the lake with different area is analyzed. The results show as follows: (1) The nearest neighbor method is more reasonable than the resampling method when the multi-scale remote sensing image is obtained by resampling remote sensing data with high spatial resolution for lake area extraction and lake spatial scale effect analysis. (2) The MTMF sub-pixel decomposition method can be used for remote sensing lake boundary extraction and area calculation based on spectral features of water body. However, it is easy to confuse lakes with rivers or other non-lake water bodies during boundary extraction. (3) The extraction results of the remote sensing lake area are greatly affected by the spatial resolution of the remote sensing image used. The lower the spatial resolution of the image, the larger the deviation of the lake area extraction, especially for lakes with smaller area.